Latest papers
In this research, a new semi-implicit two-phase double-point material point method is proposed, in which the soil and water phases are modelled using two distinct sets of material points, both being stabilised with a novel approach. The…
Stochastic world models are usually evaluated by the accuracy and calibration of their predicted futures. These criteria leave a decision-relevant ambiguity: the same conditional future distribution can arise because an observation aliases…
We propose a novel UV-IR mechanism in quantum gravity in which black hole instabilities act as a bridge to the most ultraviolet sector of the theory. Specifically, we argue that when black holes reach a critical temperature associated with…
Exploring axion-like particle (ALP) signatures from neutron stars (NSs) in the \emph{Fermi}-LAT energy range remains largely unexplored. Neutron stars with exceptionally strong magnetic fields, such as magnetars and pulsars with…
Layout-to-image diffusion models have achieved impressive semantic controllability by conditioning generation on category-level segmentation maps. However, such category-aligned control is not necessarily instance-addressable: multiple…
In this article, we introduce the notion of connected finite graphs with disjoint cycles in normal form and show that any such graph can be transformed into a normal form graph via a finite sequence of in-splittings and out-splittings.…
Despite the widespread adoption of foundation models as feature extractors for medical imaging, relatively little is understood about how different pretraining strategies influence the transferability of learned representations to weakly…
Verification for retrieval-augmented generation usually scores each retrieved chunk and drops the ones that fail. We show this cannot work for multi-hop questions, and show what does. Per-chunk scoring assumes one chunk is a sufficient…
Recent advances in image generation and editing have made prompt quality a key bottleneck for e-commerce creatives. Vision-language models (VLMs) can generate image-editing prompts from product images and metadata, but further improving…
Chain-of-thought (CoT) monitoring is meant to catch the reward hacks that look clean in the actions and betray themselves only in the reasoning. We show that this is exactly where an adversary who controls the reasoning can defeat it.…
Pretrained byte-level BPE tokenizers can segment underrepresented languages inefficiently. Replacing a tokenizer changes the meaning of nearly every token ID, while vocabulary expansion enlarges the model's embedding and output matrices. We…
Off-the-shelf LID and letter heuristics over-label Kazakh-Russian social text as mixed: Russian loanwords inside Kazakh look like code-switching under a shared Cyrillic script. We release a document-level gold LID set whose guideline keeps…
We revisit a central estimate in the economics of education: the human-capital loss associated with COVID-19 school closures. Estimates of pandemic learning loss may be affected by publication bias, p-hacking, and the mechanical correlation…
We bridge two sides of singular perturbation theory: the classical theory of slow-fast systems and the semi-classical approach to quantum mechanical systems. For a specific but physically important class of dynamical systems, we show that…
A new semi-implicit, two-phase, double-point formulation of the Material Point Method (MPM) for soil-water interaction with seepage and free-surface flows under large deformation is presented in this paper. The approach advances the water…
Mixture-of-Experts (MoE) models have become a dominant architecture for large-scale AI services, yet deploying them over geo-distributed heterogeneous edge servers remains challenging. When the Top-k activated experts of a token are spread…
High-polyphony symbolic music is increasingly used in generation, analysis, and arrangement, yet many downstream tasks require bounded representations with fixed tracks or slots. Converting richly orchestrated scores into compact forms is…
Subbarao and Verma introduced, in 1999, a number of open problems concerning the sequence $(f(n))_{n \geq 0}$ of complementary Bell numbers, which may be defined via Bell polynomials $B_{n}(x) = \sum_{k=0}^{n} \left\{ \begin{smallmatrix} n…
Vision-language MoE batches contain different numbers of image and text tokens. Image resolution, image count, tiling, and prompt length all change this token mix. We call the standard token-level Switch auxiliary loss Std-Aux. Std-Aux…
Serving Mixture-of-Experts (MoE) large language models across distributed edge servers is bottlenecked by the cross-server expert transmission. The existing approaches mainly focus on how to reach a remote expert faster. However, in this…